Slideshare uses cookies to improve functionality and performance, and to provide you with relevant advertising. If you continue browsing the site, you agree to the use of cookies on this website. See our User Agreement and Privacy Policy.

Slideshare uses cookies to improve functionality and performance, and to provide you with relevant advertising. If you continue browsing the site, you agree to the use of cookies on this website. See our Privacy Policy and User Agreement for details.

Like this presentation? Why not share!

- Data Mining with Excel 2010 and Pow... by Mark Tabladillo 14858 views
- Excel Datamining Addin Beginner by excel content 6629 views
- Analysis with Excel CUBE Functions ... by PowerBIUKCommunity 357 views
- Bay Area Microsoft Business Intelli... by Mark Ginnebaugh 1959 views
- Data Mining 2012 with Excel and Pow... by Boriana Petrova 1395 views
- Data analysis using excel by aazoman 54900 views

4,722 views

4,336 views

4,336 views

Published on

This presentation is for those looking to drive more from Excel when reporting from the BI Semantic Model, whether it be a PowerPivot data model, a tabular database or multidimensional cube.

This is an alternative approach to reporting from Analysis Services data models that provides control and customization opportunities not (easily) supported by conventional data model reporting tools such as PivotTable, Power View, Reporting Services, etc.

You'll Learn

• About each of the seven CUBE functions

• Practical uses and considerations for building creative reports and dashboards.

• How to create dynamic parameters to produce dynamic reporting content

• When to use, and when not to use, the CUBE functions

• How to describe common and advanced scenarios for using the CUBE functions in report designs

Peter Myers specializes in Microsoft Business Intelligence (BI) and provides mentoring, technical training and course content authoring for Microsoft SQL Server and Microsoft Office. Specifically, he works with SQL Server BI and Office BI, including Power BI.

No Downloads

Total views

4,722

On SlideShare

0

From Embeds

0

Number of Embeds

19

Shares

0

Downloads

3

Comments

0

Likes

6

No embeds

No notes for slide

- 1. Mastering the Excel CUBE Functions Bay Area BI User Group | Mountain View | 8 May, 2014
- 2. Presenter Introduction Peter MyersPeter MyersPeter MyersPeter Myers BI Expert – Bitwise Solutions BBus, SQL Server MCSE, MCT, SQL Server MVP (since 2007) Experienced in designing, developing and maintaining Microsoft database and application solutions, since 1997 Focuses on education and mentoring Based in Melbourne, Australia peter.myers@bitwisesolutions.com.au http://www.linkedin.com/in/peterjsmyers
- 3. Session Objectives To introduce each the Excel CUBE functions To describe when to use, and not to use, the CUBE functions To provide recommended practices To suggest advanced scenarios To demonstrate how to create compelling and interactive reports based on the CUBE functions
- 4. Session Outline Introducing the CUBE Functions When to Use the CUBE Functions When Not to Use the CUBE Functions Recommended Practices Advanced Scenarios Demonstrations
- 5. Introducing the CUBE Functions Excel 2013 includes seven CUBE functions – available since Excel 2007 CUBE functions allow creating Excel report layouts cell-by- cell based on an Analysis Services BI Semantic Model (BISM) Only works for data models that can be queried with Multidimensional Expressions (MDX) Tabular data models in DirectQuery mode cannot be queried by using MDX Reports can allow user input by using slicers, timelines and cell values (simulating parameters) CUBE function arguments can be defined by using expressions – both by using Excel or MDX expressions
- 6. Introducing the CUBE Functions Data model metadata, members, sets, tuples and formats are cached to minimize impact on the database Using the CUBE functions can work well for some report designs, and not so well for others CUBE function report layouts can be created by converting a PivotTable to formulas – it is a one-way conversion, and can be helpful for beginners learning to write CUBE formulas Continued
- 7. Demonstrations Converting a PivotTable to Formulas
- 8. Introducing the CUBE Functions Excel includes seven CUBE functions: CUBEMEMBER CUBESET CUBESETCOUNT CUBERANKEDMEMBER CUBEVALUE CUBEKPIMEMBER CUBEMEMBERPROPERTY Continued
- 9. Introducing the CUBE Functions Most CUBE functions require Connection and Caption parameters Connection:Connection:Connection:Connection: Text string of the name of the workbook connection to the data model Create workbook connections to Analysis Services data models, or use the PowerPivot data model workbook connection “ThisWorkbookDataModel” Tip: Name the workbook connections appropriately – and concisely – by using the Friendly Name property Caption:Caption:Caption:Caption: Text string that is displayed in the cell instead of the default caption, if one is defined, from the data model Use this argument to present a user-friendly label Continued
- 10. Introducing the CUBE Functions Returns a member or tuple from the data model member_expressionmember_expressionmember_expressionmember_expression:::: Text string of an MDX expression that evaluates to a unique member in the data model. Alternatively, it can be a tuple, specified as a cell range or an array constant. caption:caption:caption:caption: When a tuple is returned, the caption used is the one for the last member in the tuple Examples of member_expressionmember_expressionmember_expressionmember_expression inputs: [Date].[Calendar].[CY2013] A$1 (A2, B2, C2) A2:C2 CUBEMEMBER(connection, member_expression, caption)
- 11. Introducing the CUBE Functions Returns a set of members from the data model set_expressionset_expressionset_expressionset_expression:::: A set expression sort_ordersort_ordersort_ordersort_order:::: Integer value specifying the rank to return sort_bysort_bysort_bysort_by:::: Text string of the value by which to sort, typically a measure Note: The sort parameters are ideal for passing in user preference (parameters) Excel slicers (representing multi-member selection) can be passed in to the set_expressionset_expressionset_expressionset_expression argument CUBESET(connection, set_expression, caption, sort_order, sort_by)
- 12. Introducing the CUBE Functions Returns the number of items in a set set:set:set:set: Text string or an Excel expression that evaluates to a set defined by the CUBESET function Note: This is the only CUBE function that does not require a connection It does require the result of a CUBESET function to be passed in Can be used to determine the number of members selected in an Excel slicer Can be used to determine report layout (to simulate dynamic rows/columns) CUBESETCOUNT(set)
- 13. Introducing the CUBE Functions Returns the nth, or ranked, member in a set set_expressionset_expressionset_expressionset_expression:::: A set expression, a CUBESET function, or reference to a CUBESET function rank:rank:rank:rank: Integer value specifying the rank to return Note: The first member is rank 1 Requesting a rank beyond the set member count will produce an error Can work around this by using IFERROR(CUBERANKEDMEMBER(…), 0) To return the bottom n values, use the sort_ordersort_ordersort_ordersort_order and sort_bysort_bysort_bysort_by arguments of the CUBESET function to reverse the order of the set Works best when cube sets have a fixed, and known, number of members, such as the top sales performer, the top 10 students, or the 12 months of a year Use the ROW or COLUMN function to build a “template function” to copy and paste into a range CUBERANKEDMEMBER(connection, set_expression, rank, caption)
- 14. Introducing the CUBE Functions Returns an aggregated value from the data model member_expressionmember_expressionmember_expressionmember_expression:::: Text string of an MDX expression that evaluates to a member or tuple within the data model. Alternatively, it can be a set defined with the CUBESET function. Note: Each member_expression will form a tuple If at least one element within the tuple is invalid, CUBEVALUE returns a #VALUE! error value Excel slicers and timelines can be used as member expressions CUBEVALUE(connection, member_expression1, member_expression2…)
- 15. Introducing the CUBE Functions Returns a KPI property from the data model kpi_namekpi_namekpi_namekpi_name:::: Text string of the name of the KPI in the data model kpi_propertykpi_propertykpi_propertykpi_property:::: Integer value specifying the KPI component to be returned, and can be one of the following: CUBEKPIMEMBER(connection, kpi_name, kpi_property, caption) Integer Enumerated constant 1 KPIValue 2 KPIGoal 3 KPIStatus 4 KPITrend 5 KPIWeight 6 KPICurrentTimeMember Note: Status and Trend values represent a normalized value, typically: Status: -1 (off track), 0 (slightly off track) and 1 (on track) Trend: -1 (deteriorating), 0 (no change) and 1 (improving) Auto-complete does not list the available status or trend values – you must know them – consult the data model documentation ☺ Use Excel conditional formatting to hide values and represent them as color scales or icon sets
- 16. Introducing the CUBE Functions Returns the value of a member property from a cube member_expressionmember_expressionmember_expressionmember_expression:::: Text string of an MDX expression of a member within the cube, or a reference to an Excel formula that returns a member (CUBEMEMBER or CUBERANKEDMEMBER) property:property:property:property: Text string of the name of the property returned, or a reference to a cell that contains the name of the property Note: Useful for introducing properties like phone number, email address, price, etc. Auto-complete does not list the available member properties – you must know them – consult the data model documentation ☺ Tabular BISM, including Power Pivot, cannot define member properties CUBEMEMBERPROPERTY(connection, member_expression, property)
- 17. Demonstrations Producing a Dashboard with the CUBE Functions
- 18. When to Use the CUBE Functions When dashboard-style layouts are required, especially when published to SharePoint and rendered by Excel Services To reproduce reports that must look and behave similarly to legacy reports To overcome limitations of PivotTables and PivotCharts Create asymmetric reports (this year vs. last year) – though this could possibly be achieved with a custom set GETPIVOTDATA is not an elegant or easy way to retrieve values from a PivotTable When customized MDX is required To integrate data from multiple data models To produce easier logic by using Excel formula, instead of MDX or DAX in the data model
- 19. When Not to Use the CUBE Functions When a PivotTable or PivotChart, or Power View, could deliver the same result These report types provide structured exploration and rich interactivity These report types are easier to create When interactivity is required: sorting, filtering, etc. Some of these can be simulated with CUBE formulas by using expressions, but this takes more development time and effort When sets with variable member counts need to be displayed – the CUBERANKEDMEMBER function will return an error if requesting a rank beyond the set size You can work around this by testing with the IFERROR function
- 20. Recommendations Learn the CUBE functions by starting with a PivotTable and converting it to formulas Know MDX to customize beyond the basics – you can work beyond the Auto-complete options Adopt a methodical design approach to help implement change and to troubleshoot errors Create a “Globals” worksheet to define CUBE formulas and annotate them Hide rows, columns or worksheets that contain CUBE formulas Avoid complex CUBE formulas – break them down to smaller formulas and nest them Strive for parameter-driven designs to allow for variability and user interactivity Define Excel named ranges to simplify building member and set expressions Use “dynamic” expressions based on Excel formula Use slicers and timelines to allow users to interactively filter the report
- 21. Recommendations Use template functions to copy and paste into ranges This is especially useful for the CUBERANKEDMEMBER and CUBEVALUE functions Define captions to allow the CUBE formulas to blend into the report If the data model definition changes, be sure to refresh the workbook connection to update the cache This is usually important when modifying the embedded PowerPivot data model Expressions cannot be greater than 255 characters, which is the limit for an argument to a function in Excel To overcome the text length limit, use references to cells that contain the expression, which can have a text length size of up to 32,767 characters Continued
- 22. Advanced Scenarios Produce dynamic layouts by using macros However, these reports cannot be rendered by Excel Services Produce reports that integrate members and values from multiple data models Combine the enterprise data model with your PowerPivot data model Extend the corporate data stores with custom data and/or calculations Tip: It pays off to design data models with a consistent structure (dimension, hierarchy, member names, etc.) – but this is not necessary, just easier Use the Table Analysis Tools Excel data mining add-in based on an Excel table of CUBE formulas Create an Excel table based on CUBE formulas as a source for other visualizations: Power Map Visio Services Apps for Office
- 23. Demonstrations Advanced Scenarios with the CUBE Functions
- 24. Resources Excel CUBE Functions wiki (coming soon) http://www.excelcubefunctions.com Channel 9: TechEd New Zealand 2013 Video: “Mastering the Excel CUBE Functions in Excel 2013” http://channel9.msdn.com/Events/TechEd/NewZealand/2013/DBI304
- 25. Thank You

No public clipboards found for this slide

Be the first to comment